This document is part instruction manual and part data summary tool focused on the collection, wranggling, and summarizing passive audio recordings (PAR) data. Data are generated by automated recording units (ARUs) deployed along an urban rural gradient from Newark, DE to Landenberg, PA.
The first step in this process is to deploy the ARU’s at the appropriate sites (See map and table below) with fresh batteries and formatted SD Cards. We have found that the AudioMoth ARUs will last 1.5 months when they have fresh AA batteries and are recording continuously (10 minute file saves) for two hours in the morning and two hours in the evening.
We selected Carolina Wren song vocalizations to train the BirdNET model and screen the XX months of recordings. Carolina Wrens vocalization types used in this analysis include;
We located 20 Automated Recording Units (ARUs) along an urban / rural gradient, like so many others have done before us! We placed ARUs at locations where Carolina Wrens were detected on at least two of three visits with the motivation that our primary objective was to describe the phenological variation in Carolina Wren vocalizations over the annual cycle.
| Site_ID | latitude | longitude |
|---|---|---|
| 1 | 39.76574 | -75.78674 |
| 2 | 39.75722 | -75.78384 |
| 3 | 39.75202 | -75.77538 |
| 4 | 39.74589 | -75.77399 |
| 5 | 39.73643 | -75.76424 |
| 6 | 39.71957 | -75.76209 |
| 7 | 39.71193 | -75.75936 |
| 8 | 39.71631 | -75.76090 |
| 9 | 39.72483 | -75.76762 |
| 10 | 39.72924 | -75.76563 |
| 11 | 39.68237 | -75.74138 |
| 12 | 39.69333 | -75.75937 |
| 13 | 39.69598 | -75.75568 |
| 14 | 39.68959 | -75.75491 |
| 15 | 39.68782 | -75.74917 |
| 16 | 39.69389 | -75.76287 |
| 17 | 39.66918 | -75.75050 |
| 18 | 39.67336 | -75.74437 |
| 19 | 39.67873 | -75.74261 |
| 20 | 39.67142 | -75.74995 |
## ADD lapply code for making am pm figures here...
song.dat <- read.csv('../Reality-Wren/Output/wren_detections.csv')
## Create a column for morning (am) and evening (pm)
song.dat <- song.dat %>%
mutate(time_code = case_when(time < 1200 ~ 'am',
time > 1200 ~ 'pm'))
## Create df grouped by time_code (am & pm)
total_wren_day <- song.dat %>%
group_by(site, month, day, time_code) %>%
summarize(total_wren = sum(count))
total_wren_day$site <- as.factor(total_wren_day$site)
sites <- levels(total_wren_day$site)
## create a plot function that will use lapply to iterate through sites making a figure for each site (like above but this time with stacked bars showing songs in am and pm)
## Get file names to change
my_plot_hours <- function(sites){
total_wren_day %>%
filter(site == sites) %>%
ggplot(aes(fill=time_code, y=total_wren, x=day)) +
geom_bar(position="stack", stat="identity")+
theme_classic() +
labs(
x = "day",
y = "total wren detections",
title = paste("Reality Wren Phenology Mar-Oct Site", sites)) +
facet_grid(rows=vars(month))
}
lapply(sites, my_plot_hours)
## [[1]]
##
## [[2]]
##
## [[3]]
##
## [[4]]
##
## [[5]]
##
## [[6]]
##
## [[7]]
##
## [[8]]
##
## [[9]]
##
## [[10]]
##
## [[11]]
##
## [[12]]
##
## [[13]]
##
## [[14]]
##
## [[15]]
##
## [[16]]
##
## [[17]]
##
## [[18]]
##
## [[19]]
##
## [[20]]
## run the vector of sites through the function using lapply
#htmltools::tagList(lapply(sites, my_plot_hours))